Question about iteration

[YunfeiWang]

Dear all

       If I wrote an iteration method in my  umatln  file , How can I use the solution command of LOOP...NEXT?
   
      In my opinion,  I should  run the programs with the command LOOP time n....NEXT and without LOOP iter ... NEXT?  So the solution command

 in my input file is:
 
                                                                                     DT,,i
                                                                                     LOOP time j
                                                                                     TIME
                                                                                              tang,,1
                                                                                     NEXT

      Could you tell me whether I thought is correct ?  If not, how could I modify it？
           
Thank you very much

Best regards   

[FEAP_Admin]

What is a umatin file?

Note, FEAP's LOOP - NEXT structure will repeat what ever commands you place in the body of the loop.

[YunfeiWang]

What is a umatin file?

Note, FEAP's LOOP - NEXT structure will repeat what ever commands you place in the body of the loop.

Umatln is an user material model

[Prof. R.L. Taylor]

Can you explain why you need to use a loop command when you are doing input for a user material model? 

Or are you asking how to program the input module to read data using a do-while loop in the umati*.f file?

As written your question does not make sense to us.

[YunfeiWang]

Can you explain why you need to use a loop command when you are doing input for a user material model? 

Or are you asking how to program the input module to read data using a do-while loop in the umati*.f file?

As written your question does not make sense to us.

Dear Prof

       The reason for using a loop command is that I want use a proporitional load.  My problem is non-liner, but I has written the Newton-Rhapson  iteration  in

 my user material model. So I want to ask that can I compile my input file with the loop command without  iteration. So the format would be :
 
                 DT,,dt-value                             
                 LOOP,,n-steps                             
                    TIME                                           
                    TANG,,1               
                  NEXT       

Thank you for your answer

Best Regards   

[JStorm]

You need to  define an incremental material formulation and to use the newton-raphson loop of feap together with proportional boundary conditions. I'm still not sure about your intention.
or do you define a material formulation with an internal iteration in order to stabilise the convergence of the material response?

[YunfeiWang]

You need to  define an incremental material formulation and to use the newton-raphson loop of feap together with proportional boundary conditions. I'm still not sure about your intention.
or do you define a material formulation with an internal iteration in order to stabilise the convergence of the material response?

Dear JStorm

       As you said, I have define a material formulation(umatln and n=1~9) with an internal iteration. So I want to ask does it  need to edit an

iteration command in my input file?

[FEAP_Admin]

LOOP,,10
  TANG,,1
NEXT

Will solve the GLOBAL equations with a Newton-Raphson scheme.  At each iteration, your material model will be called.  If your material model needs to perform a LOCAL Newton-Rahpson iteration to determine its state that is perfectly fine.  Just program the LOCAL Newton-Raphson loop inside your material model.  FEAP will make sure to call your material for each GLOBAL iteration so that your user material can compute the LOCAL iterations.

[Prof. R.L. Taylor]

When you solve any problem for the first times it is always good to iterate on the solution part, that is, for a problem with a proportional load use

LOOP,,nt   (where nt is the number of time steps)
  TIME
   LOOP,,5
      TANG,,1
  NEXT,ni
    ....
NEXT,nt

If the problem converges in one iteration then you can replace the "5" by a "1" -- but it is good to leave the loop so the "jump" is to the command after the NEXT,ni

[YunfeiWang]

LOOP,,10
  TANG,,1
NEXT

Will solve the GLOBAL equations with a Newton-Raphson scheme.  At each iteration, your material model will be called.  If your material model needs to perform a LOCAL Newton-Rahpson iteration to determine its state that is perfectly fine.  Just program the LOCAL Newton-Raphson loop inside your material model.  FEAP will make sure to call your material for each GLOBAL iteration so that your user material can compute the LOCAL iterations.

Dear Admin

       Thank you very much. Its a very useful answer

Best regards

[YunfeiWang]

When you solve any problem for the first times it is always good to iterate on the solution part, that is, for a problem with a proportional load use

LOOP,,nt   (where nt is the number of time steps)
  TIME
   LOOP,,5
      TANG,,1
  NEXT,ni
    ....
NEXT,nt

If the problem converges in one iteration then you can replace the "5" by a "1" -- but it is good to leave the loop so the "jump" is to the command after the NEXT,ni

Dear Prof

       Thank you for your  patient explaining。 Its help me so much.

Best Regards